Modulation of blood-brain barrier glucose transport is postulated to be of clinical importance in insulin dependent diabetes mellitus (IDDM)). When plasma glucose concentrations fall, patients with poorly controlled IDDM experience symptoms and activation of counterregulatory hormones at higher plasma glucose levels than do non-diabetic subject. A single two-hour episode of hypoglycemia the previous afternoon in patients with IDDM will lower the plasma glucose level at which symptoms and hormonal counterregulation occur the next morning. Hypothesis; Changes in behavioral and counterregulatory hormonal responses to hypoglycemia that occur in subjects with IDDM are associated with reduced blood-to-brain glucose transport (Jin) which normalizes after antecedent hypoglycemia. Specific Aim IA: Measure Jin with positron emission tomography (PET) and 1-11C-D-glucose (1-11CG) during hyperinsulinemic hypoglycemic clamp (3.6 muml 1) in patients with poorly controlled IDDM and in nondiabetic subjects. Specific Aim IB: Measure J in with PET and 1-11 CG during hyperinsulinemic hypoglycemic clamp (2.8 mumol ml 1) in patients with IDDM on two occasions, once following clamped hypoglycemia (2.8mumol ml 1) in patients with IDDM on two occasions, once following clamped hypoglycemia (2.8 mumol ml 1) the previous afternoon and once following otherwise identical clamped hyperglycemia (11.1 mumol mml1) the previous afternoon. A better understanding of the adaptive capacity of BBB glucose transport and its relationship to warning symptoms of hypoglycemia and the hormonal counterregulatory response would be of great value in designing better therapeutic strategies to minimize the deleterious effects of iatrogenic hypoglycemia. Under conditions of increased neuronal activity, local CMRG1u increases markedly. Data regarding concomitant changes in J in are inconsistent. PET with 1-11CG has demonstrated accuracy for measurements in resting brain during local physiologic brain activation when production of 11C=-lactate may increase is unknown. Hypothesis; PET with 1-11CG provides accurate measurement of CMRGlu during local physiologic brain activation. Specific Aim II: Perform paired PE T measurements of CMRGlu with 1-11CG and 18F- fluorodeoxyglucose (18FDG) during visual stimulation on successive days in normal subjects. The ability to measure j in and CMRGlu with 1-11CG during physiologic brain activation under normal conditions and in abnormal states would provide valuable information regarding the relationship between glucose supply and demand during a variety of physiological and pathological conditions in humans.